154
U. Drescher
connections are formed has been a longstanding puzzle. The presently
prevailing hypothesis, formulated by Sperry decades ago, is the
chemoaffinity hypothesis (Sperry 1963). Sperry proposed that molecules expressed in gradients in the tectum would provide positional
information along which ingrowing retinal axons, endowed with receptors expressed in gradients, might find their correct retinotopic position.
Both EphrinA2 and EphrinA5 are expressed in overlapping highcaudal-to-Iow-rostral gradients in the tectum (Monschau et al. 1997),
and a corresponding receptor, EphA3, is found in a high-temporal-tolow-nasal expression in the retina (Cheng et al. 1995). Their complementary patterns make these molecules likely candidates for an involvement in the topographic mapping of retinal axons as further supported
by comparatively high binding affinities between EphA3 and
EphrinA2/EphrinA5 (Monschau et al. 1997). Besides these three molecules there are at least nine other Eph family members expressed during
development of the retinotectal projection, often in dynamic spatially
and temporally restricted patterns (for review see Drescher et al. 1997;
Pasquale 1997; Flanagan and Vanderhaeghen 1998; O'Leary and
Wilkinson 1999).
In vitro, EphrinA5 repels both nasal and temporal axons, the latter
being more sensitive than the former (Monschau et al. 1997). EphrinA2
exhibits a repellent activity exclusively for temporal axons, with no
effect on nasal axons (Nakamoto et al. 1996; Monschau et al. 1997). In
vivo, temporal but not nasal axons avoid patches of ectopically expressed EphrinA2 (Nakamoto et al. 1996). A loss-of-function analysis
showed that EphrinA5 is essential for the correct topographic mapping
of retinal axons in the mammalian visual system (Frisen et al. 1998).
Here, temporal axons formed termination zones not only in the rostral,
but also in the topographically inappropriate caudal superior colliculus,
a structure homologous to the chick optic tectum. Additionally, a transient overshooting of retinal axons into the inferior colliculi in these
mutant mice points to a function of EphrinA5 in limiting the growth of
retinal axons to the superior colliculi.
Interestingly, EphrinA2 and EphrinA5 expression is seen not only in
the tectum, but also in the retina (Marcus et al. 1996; Brennan et al.
1997; Connor et al. 1998), where they are expressed in a high-nasal-tolow-temporal pattern on retinal ganglion cell axons, thus colocalizing
with EphA receptors (Fig. 1; Hornberger et al. 1999). In the stripe assay,
U. Drescher
connections are formed has been a longstanding puzzle. The presently
prevailing hypothesis, formulated by Sperry decades ago, is the
chemoaffinity hypothesis (Sperry 1963). Sperry proposed that molecules expressed in gradients in the tectum would provide positional
information along which ingrowing retinal axons, endowed with receptors expressed in gradients, might find their correct retinotopic position.
Both EphrinA2 and EphrinA5 are expressed in overlapping highcaudal-to-Iow-rostral gradients in the tectum (Monschau et al. 1997),
and a corresponding receptor, EphA3, is found in a high-temporal-tolow-nasal expression in the retina (Cheng et al. 1995). Their complementary patterns make these molecules likely candidates for an involvement in the topographic mapping of retinal axons as further supported
by comparatively high binding affinities between EphA3 and
EphrinA2/EphrinA5 (Monschau et al. 1997). Besides these three molecules there are at least nine other Eph family members expressed during
development of the retinotectal projection, often in dynamic spatially
and temporally restricted patterns (for review see Drescher et al. 1997;
Pasquale 1997; Flanagan and Vanderhaeghen 1998; O'Leary and
Wilkinson 1999).
In vitro, EphrinA5 repels both nasal and temporal axons, the latter
being more sensitive than the former (Monschau et al. 1997). EphrinA2
exhibits a repellent activity exclusively for temporal axons, with no
effect on nasal axons (Nakamoto et al. 1996; Monschau et al. 1997). In
vivo, temporal but not nasal axons avoid patches of ectopically expressed EphrinA2 (Nakamoto et al. 1996). A loss-of-function analysis
showed that EphrinA5 is essential for the correct topographic mapping
of retinal axons in the mammalian visual system (Frisen et al. 1998).
Here, temporal axons formed termination zones not only in the rostral,
but also in the topographically inappropriate caudal superior colliculus,
a structure homologous to the chick optic tectum. Additionally, a transient overshooting of retinal axons into the inferior colliculi in these
mutant mice points to a function of EphrinA5 in limiting the growth of
retinal axons to the superior colliculi.
Interestingly, EphrinA2 and EphrinA5 expression is seen not only in
the tectum, but also in the retina (Marcus et al. 1996; Brennan et al.
1997; Connor et al. 1998), where they are expressed in a high-nasal-tolow-temporal pattern on retinal ganglion cell axons, thus colocalizing
with EphA receptors (Fig. 1; Hornberger et al. 1999). In the stripe assay,